The statements in this section merely provide background information related to the present disclosure and may not constitute prior art.
A load may appear to a power supply as a resistive impedance, an inductive impedance, a capacitive impedance, or a combination thereof. When the current passing to the load is in phase with the voltage applied to the load, the power factor approaches one.
When the power factor is less than one, transmitted power can be wasted (due to phase mismatch between current and voltage) and/or noise may be introduced into the power line. To reduce noise and improve efficiency, power supplies generally use power factor correction (PFC) circuits to control the phase of the current waveform relative to the phase of the voltage waveform.
Referring now to FIG. 1, a conventional boost converter 10 includes rectifier 15, which receives alternating current (AC) power. Input current Iin passes through inductor 20 and part of input current Iin passes through diode 50 (having a capacitor/filter 60 at its output) before being applied to load 70.
Power factor controller 30 controls the current flowing through inductor 20 by turning switch 40 on and off in response to an AC voltage-sensing input 12, a DC output voltage 72, a sensed power conversion current from a second inductor coil 25, and a feedback current via node 34. When switch 40 is on, current 22 generally flows through inductor 20 (thereby storing some energy in inductor 20) and then through switch 40 to ground. When switch 40 is off, current 52 may flow through diode 50 and some charge may collect on capacitor/filter 60. Generally, current flow 22 through inductor 20 is significantly reduced or even prevented when the switch 40 is off.
Referring now to FIG. 2, an AC voltage V received by boost converter 10 is shown. Input voltage V is a rectified half-sine wave of the AC waveform input. However, due to the on/off cycles of switch 40 (controlled by power factor controller 30 in FIG. 1), the current waveform I in FIG. 2 has a sawtooth pattern. After passing the sawtooth waveform I through a low-pass filter (e.g., high frequency bypass capacitor/filter 60 in FIG. 1), the input current waveform resembles the input AC voltage at the input of rectifier 15.
The PF for the conversion approaches 1 under most conditions, particularly those conditions where the loading power is sufficiently high to allow an appreciable average input current to continuously pass through inductor 20. This mode is known as the “average current mode” or “continuous mode” of operation for boost converter 10.
The PFC for a boost converter generally has two parameters defined by a specification: (1) PF, and (2) total harmonic distortion (or THD). THD refers to distortion caused generally by higher order harmonics. For a 60 Hertz (Hz) AC signal, higher order harmonics are located at 120 Hz, 180 Hz, or other n*60 Hz values, where n is an integer of 2 or more. Generally, the higher the THD, the lower the efficiency. Harmonic distortion can saturate inductor 20 in boost converter 10. Moreover, if the THD is sufficiently high, noise can be fed back onto the AC power lines 12-14, which is undesirable.
Referring now to FIG. 3A, a low-power and/or low-voltage portion 120 of the voltage and current waveforms of FIG. 2 are shown. The voltage waveform V is the voltage at the output of rectifier 15 (see FIG. 1). The current waveform I is the input current Iin passing through inductor 20. When switch 40 in FIG. 1 is turned on at time t0, current I increases in a substantially linear manner, as shown by slope 122. Switch 40 is on for a period of time determined by power factor controller 30. At the end of this time (point 124 on the current waveform I in FIG. 3A), switch 40 turns off and current I decreases in a substantially linear manner. Switch 40 then is turned on again by power factor controller 30 (see FIG. 1) after a period of time ts-t0, also determined by power factor controller 30.
When current I=0 (i.e., I0, the current value during “zero current period” 126 in FIG. 3A), the average current or continuous mode of operation has a potential distortion issue. The THD cannot be controlled during the zero current period 126 of waveform portion 120 because there is no current flowing through inductor 20 of FIG. 1.
The discontinuous mode of operation of boost converter 10 occurs during periods of time where switch 40 is turned on and off for lengths of time sufficient for zero current periods to occur. The critical mode of operation occurs when current waveform I (see FIG. 3A) is at or near zero (I0). It is desirable to maximize the amount of time that the inductor current Iin is above zero (see FIG. 1) and to minimize the zero current periods (e.g., zero current period 126 of FIG. 3A).
Referring now to FIG. 3B, ideally ts would occur at a point in time when current I crosses I0 (the “I=0” axis), zero current period 126 would have a duration as close to 0 units of time as possible, and switch 40 (see FIG. 1) would be turned on essentially immediately by power factor controller 30 (see FIG. 1) after current waveform portion 134 intersects I0 (see FIG. 3B). When this occurs, current waveform portion 136 increases soon after current waveform portion 134 intersects I0 and current to flow through inductor 20 (see FIG. 1) substantially continuously. The switch 40 should not be turned on too soon (i.e., before current waveform portion 134 in FIG. 3B intersects I0). When this occurs, the average input current may increase at too high a rate, which could cause the input current waveform phase to move out of alignment with the input voltage waveform phase.
One conventional approach detects the input current Iin flowing through inductor 20 in FIG. 1. A second inductor coil 25 magnetically coupled to the inductor 20 senses the current Iin flowing through inductor 20. However, this approach suffers from latency when sensing the current in another coil. The latency introduces some positive length of time in the zero current period 126 (see FIG. 3A) and noise back into the AC power line 12-14. Also, the second inductor coil 25 adds some expense to manufacturing power factor controller 30 and necessitates at least one dedicated differential pin on power factor controller 30 to receive information from second inductor coil 25.
Another approach attempts to sense the current at node 34 in FIG. 1. However, the current and voltage values at node 34 are relatively low in the critical mode of operation. As a result, error signals based on the measurement are relatively inaccurate. Also, determining the current at node 34 would require power factor controller 30 to have a relatively high sampling rate (i.e., >>1 sample taken every 1/[ts−t0] seconds) in the critical mode, and the sampling resolution should be relatively high to avoid turning switch 40 on too fast or too slow.